Kubernetes 1.17 containerd 1.2.0 与 Calico CNI 节点未加入主节点

Question

我正在使用 containerd 和 Calico 作为 CNI 在 CentOS 8 上设置 kubernetes 集群。使用 kubeadm 命令设置主节点，其处于就绪状态。

当我将节点加入主节点时，节点未进入就绪状态。我看到下面的日志文件消息。

Jan 14 20:17:29 node02 containerd[1417]: time="2020-01-14T20:17:29.416373526-05:00" level=error msg="RunPodSandbox for &PodSandboxMetadata{Name:calico-node-fbst8,Uid:9c7f6334-d106-48e1-af12-1bcdebc7c2c2,Namespace:kube-system,Attempt:0,} failed, error" error="failed to start sandbox container: failed to create containerd task: OCI runtime create failed: container_linux.go:348: starting container process caused "process_linux.go:279: applying cgroup configuration for process caused \"Invalid unit name 'pod9c7f6334-d106-48e1-af12-1bcdebc7c2c2'\"": unknown"
Jan 14 20:17:29 node02 kubelet[30113]: E0114 20:17:29.416668   30113 remote_runtime.go:105] RunPodSandbox from runtime service failed: rpc error: code = Unknown desc = failed to start sandbox container: failed to create containerd task: OCI runtime create failed: container_linux.go:348: starting container process caused "process_linux.go:279: applying cgroup configuration for process caused \"Invalid unit name 'pod9c7f6334-d106-48e1-af12-1bcdebc7c2c2'\"": unknown
Jan 14 20:17:29 node02 kubelet[30113]: E0114 20:17:29.416742   30113 kuberuntime_sandbox.go:68] CreatePodSandbox for pod "calico-node-fbst8_kube-system(9c7f6334-d106-48e1-af12-1bcdebc7c2c2)" failed: rpc error: code = Unknown desc = failed to start sandbox container: failed to create containerd task: OCI runtime create failed: container_linux.go:348: starting container process caused "process_linux.go:279: applying cgroup configuration for process caused \"Invalid unit name 'pod9c7f6334-d106-48e1-af12-1bcdebc7c2c2'\"": unknown
Jan 14 20:17:29 node02 kubelet[30113]: E0114 20:17:29.416761   30113 kuberuntime_manager.go:729] createPodSandbox for pod "calico-node-fbst8_kube-system(9c7f6334-d106-48e1-af12-1bcdebc7c2c2)" failed: rpc error: code = Unknown desc = failed to start sandbox container: failed to create containerd task: OCI runtime create failed: container_linux.go:348: starting container process caused "process_linux.go:279: applying cgroup configuration for process caused \"Invalid unit name 'pod9c7f6334-d106-48e1-af12-1bcdebc7c2c2'\"": unknown
Jan 14 20:17:29 node02 kubelet[30113]: E0114 20:17:29.416819   30113 pod_workers.go:191] Error syncing pod 9c7f6334-d106-48e1-af12-1bcdebc7c2c2 ("calico-node-fbst8_kube-system(9c7f6334-d106-48e1-af12-1bcdebc7c2c2)"), skipping: failed to "CreatePodSandbox" for "calico-node-fbst8_kube-system(9c7f6334-d106-48e1-af12-1bcdebc7c2c2)" with CreatePodSandboxError: "CreatePodSandbox for pod \"calico-node-fbst8_kube-system(9c7f6334-d106-48e1-af12-1bcdebc7c2c2)\" failed: rpc error: code = Unknown desc = failed to start sandbox container: failed to create containerd task: OCI runtime create failed: container_linux.go:348: starting container process caused \"process_linux.go:279: applying cgroup configuration for process caused \\\"Invalid unit name 'pod9c7f6334-d106-48e1-af12-1bcdebc7c2c2'\\\"\": unknown"
Jan 14 20:17:30 node02 containerd[1417]: time="2020-01-14T20:17:30.541254039-05:00" level=error msg="Failed to load cni configuration" error="cni config load failed: no network config found in /etc/cni/net.d: cni plugin not initialized: failed to load cni config"
Jan 14 20:17:30 node02 kubelet[30113]: E0114 20:17:30.541394   30113 kubelet.go:2183] Container runtime network not ready: NetworkReady=false reason:NetworkPluginNotReady message:Network plugin returns error: cni plugin not initialized
Jan 14 20:17:35 node02 containerd[1417]: time="2020-01-14T20:17:35.541792325-05:00" level=error msg="Failed to load cni configuration" error="cni config load failed: no network config found in /etc/cni/net.d: cni plugin not initialized: failed to load cni config"
Jan 14 20:17:35 node02 kubelet[30113]: E0114 20:17:35.541929   30113 kubelet.go:2183] Container runtime network not ready: NetworkReady=false reason:NetworkPluginNotReady message:Network plugin returns error: cni plugin not initialized

有解决此错误的提示吗？

Answer 1

您是否将--pod-network-cidr=192.168.0.0/16 设置为kubeadm init？
显然，你需要设置它。
https://kubernetes.io/docs/setup/production-environment/tools/kubeadm/create-cluster-kubeadm/#pod-network

Answer 2

因为您没有使用 docker，所以您需要明确地 setup cgroup 驱动程序。

要使用systemd cgroup driver，请在/etc/containerd/config.toml 和systemctl restart containerd 中设置plugins.cri.systemd_cgroup = true

你必须修改/var/lib/kubelet中的文件kubeadm-flags.env并设置cgroups驱动。

KUBELET_EXTRA_ARGS=--cgroup-driver=systemd

确保指向/etc/systemd/system/kubelet.service.d/10-kubeadm.conf中的上述文件

EnvironmentFile=-/var/lib/kubelet/kubeadm-flags.env